[BUG] Crashing on requests happening too fast

[PaulDotSH]

With the sample code

    let mut client = RpcClientBuilder::new()
        .build("http://127.0.0.1:38085")
        .unwrap();
    let mut wallet = client.wallet();
    println!("{:?}", wallet.open_wallet("test-wallet".to_string(), Some("test141414A".to_string())).await);
    
    let foo = wallet.get_balance(0, None).await;
    println!("{:?}", foo);

The following program crashes, with the error

Err(error sending request for url (http://127.0.0.1:38085/json_rpc): connection closed before message completed

Caused by:
connection closed before message completed)

However, adding a sleep between the open_wallet and get_balance requests makes it work as intended.

[PaulDotSH]

New info, this happens with any requests that are happening "too fast"

    println!("{:?}", daemon.get_block_count().await);
    println!("{:?}", daemon.get_block_count().await);
    println!("{:?}", daemon.get_block_count().await);
    println!("{:?}", daemon.get_block_count().await);
    println!("{:?}", daemon.get_block_count().await);

Has the same behavior

[PaulDotSH]

Adding std::thread::sleep(std::time::Duration::from_nanos(1)); on json_rpc_call before the request is made solves this, however it is not something that should be done...

[PaulDotSH]

@vorot93 @h4sh3d

[PaulDotSH]

Also, if you dont have this problem try running your blockchain locally and using --release

[mvntainer]

I also faced this issue, any plans to fix?

[h4sh3d]

I suspect the problem is on the node/wallet side and not from the crate. Did you try to query at the same pace the node with another lib or with curl directly to see if the problem still exists?